by Alex Mitchell, Customer Experience Lead at Privitar
Trust is critical to any business. The case for privacy rests on this fact. And as we found out at our inaugural Customer Advisory Board, retaining the trust of your own data teams is as important as the trust of your consumers.
It’s well known that businesses need consumer trust. Data paints an increasingly accurate picture of an individual’s life. Transactions are no longer ‘just’ transactions; they are signals that reveal intimate information about how we live our lives. As a consequence, consumers are trusting the businesses they transact with, in effect, themselves. Any business that betrays that should expect consequences. One thing hasn’t changed; customers can still choose to transact with someone else.
Respect for customer data builds trust. But as Privitar’s advisory board commented, so does using customer data to understand your customers, and turning that understanding into action. Models that make relevant recommendations – whether about useful products or lifestyle changes for better health – elevate your company to an integral part of customer’s lives. Using your customers’ data, and doing so respectfully, is critical in winning and maintaining your customer’s loyalty.
Too often, it’s when companies come to build those models that they encounter a seemingly intractable problem. Data owners can be cautious and unwilling to share the data for which they are responsible, even with internal teams. The data scientists, of course, need this data in order to put it to work for their customers. This is a double problem of trust; the lack of trust internally makes it hard to build models that make your customers trust you.
Any company that takes analytics seriously aspires to solve this problem. Several customers on our Advisory Board have started to build and plan new ways to support data scientists. Their ambition is to let data owners curate their datasets, specify appropriate protections and give data scientists a sense of the age and the fitness of data for what they want to do with it. The aspiration is to have data available instantly, with appropriate protections, to the people who need it.
This solves two problems of trust. The data goes out with appropriate protections, maintaining respect for the customer. The data scientist can understand the fitness of the dataset to their use case, building trust in the ecosystem the company has built. And for the data owner, they know that the data scientist respects the guard rails they put around the data to preserve customer privacy.
This new model of provisioning data also gives data scientists the benefit of consistency. Not only is that data consistently available – a huge hurdle for many companies! – but so is information about the suitability of that data for their use case. The metadata that data owners supply allows analytical teams to make upfront decisions about the data they need. The alternative, practiced far too widely, is asking for everything and seeing what works. It’s frustrating for the data science teams, who waste a lot of time working with unsuitable datasets, and it’s bad for customer privacy; the more data is available, the more information is likely to leak. Combined with reliable metadata, those who use the data and those who look after it can feel confident that they are using the right data and that customer privacy is protected.
Our Customer Advisory Board – and our customers – have huge ambitions. They want to free data for the innovators in their company, and they want to protect their customers’ privacy. Consistently available data with strong privacy protections built in is a critical aspiration for providing trustworthy data, consistently. When the data scientists trust in the data they get, they are free to create the products and models that build customer trust – and the data owners can trust that they are also protecting customer privacy.